Comparison of energy expenditure from lifestyle physical activities between patients with rheumatoid arthritis and healthy controls




Low energy expenditure is a risk for cardiovascular morbidity. The goals of this study were to compare energy expenditure between patients with rheumatoid arthritis (RA) and healthy controls.


A total of 121 RA patients and 120 healthy controls in New York City completed the Paffenbarger Physical Activity and Exercise Index at time of enrollment (1999–2000) and 1 year later to measure energy expenditure from walking, climbing stairs, and exercise/sports. Analyses were adjusted for age, sex, education, pain, social support, and depressive symptoms and were compared with recommended thresholds of energy expenditure.


Participants' mean age was 49 years, and 87% were women. Patients with RA expended fewer kilocalories per week than controls (mean ± SD 1,474 ± 1,198 versus 1,958 ± 1,940, P = 0.003), with most of this difference from less walking as opposed to high-intensity activities. Although similar percents of RA patients and controls met national recommendations for total weekly energy expenditure (56% versus 64% for the lower [≥1,000 kilocalories per week] threshold; P = 0.14, and 41% versus 48% for the higher [≥1,400 kilocalories per week] threshold; P = 0.17), fewer RA patients met the recommendations (≥700 kilocalories per week) for walking (32% versus 48%; P = 0.01). The strongest predictor of more energy expenditure at 1 year for both groups was more energy expenditure at enrollment.


Most of the difference in energy expenditure between RA patients and healthy controls was due to less walking. Given that walking is an effective and relatively safe lifestyle activity, increasing walking should be a priority to improve cardiovascular risk in RA.


Patients with rheumatoid arthritis (RA) are at a markedly increased risk for cardiovascular morbidity and mortality (1–3) with greater risk in those with more severe RA (4). Possible mechanisms include RA-related inflammation, dyslipidemia, endothelial dysfunction, insulin resistance, and effects on homocysteine concentration (5). Currently, there is some evidence that new combination and biologic therapies may attenuate the impact of RA on cardiovascular disease (6, 7). Although these possible effects are under investigation, national recommendations call for aggressive measures to mitigate the threat of RA by tight control of inflammation and other possible risks, as well as by the use of prophylactic therapy (8).

Although physical inactivity and sedentary behavior are other risks for cardiovascular disease (9), increasing exercise in RA patients is not widely embraced by patients and physicians (10, 11). In particular, exercise may be avoided because of current joint symptoms and the fear of future exacerbations (12), thereby amplifying the already elevated risk.

Contrary to previous beliefs that only vigorous sustained exercise was necessary to promote health, there is now evidence that lifestyle physical activity, described as short bouts of self-selected moderate-intensity activity incorporated into the daily routine, such as walking and climbing stairs (13, 14), also can provide cardiovascular and other health benefits (9). Fortunately, it is likely that this type of activity, under prudent guidance, may be well tolerated by many patients with RA. In an effort to promote such activities, expert musculoskeletal panels have formulated physical activity recommendations for individuals with arthritis (15). These recommendations are similar to national recommendations for the general population, which call for a total of 30 minutes of mild to moderate physical activity several days per week (16, 17). Depending on the way activities are performed, these recommendations translate to ∼1,000 to 1,400 kilocalories per week of energy expenditure (18), of which ∼700 kilocalories per week are recommended to be derived from walking (19).

Despite these more liberal recommendations, it is estimated that less than half of the US adult population currently meets these energy expenditure goals (20). Multiple barriers to physical activity have been proposed, such as older age, less education, less social support, more pain, and more depressive symptoms, as well as other demographic and psychosocial variables (21).

Currently, there is limited information about energy expenditure from lifestyle physical activity in RA patients. The goals of this observational study were to compare energy expenditure from exercise and lifestyle activities between a cohort of RA patients and controls, to assess energy expenditure according to current national recommendations, and to assess change in energy expenditure ∼1 year later. From these results, an additional goal was to identify areas for future interventions to increase physical activity in RA patients.


Participants included in this analysis were part of a larger longitudinal study comparing employment in RA patients and controls (22). RA patients were recruited during 1999 and 2000 from the medical practices of rheumatologists at the Hospital for Special Surgery and general internists at the Cornell Internal Medicine Associates in New York City. RA patients were eligible if they were 18 years of age or older, fluent in English or Spanish, met the American College of Rheumatology (formerly the American Rheumatism Association) criteria for RA (23), and were currently employed for salary. RA patients were contacted by telephone or approached when they came for office visits with their physicians, and were interviewed at that time.

Controls were a convenience sample recruited during the same time period from the Cornell Internal Medicine Associates. Controls were eligible if they were 18 years of age or older, fluent in English or Spanish, currently employed for salary, and had no chronic comorbidity as recorded in their medical charts. All controls were enrolled and interviewed when they came for routine office visits, primarily for preventive care. Controls were not matched one-to-one with RA patients, but rather were representative of patients followed in this general practice and, as a group, had similar demographic characteristics to the RA patients.


At enrollment, all participants completed a battery of questionnaires addressing employment issues and covariates, including psychosocial variables, fatigue, and physical and mental well-being as described previously (22, 24). For this analysis, we focused on measures of self-reported energy expenditure and major variables reported to be associated with physical activity, specifically certain demographic characteristics (age, sex, and education), social support, bodily pain, and depressive symptoms (25).

Energy expenditure.

Energy expenditure was recorded with the Paffenbarger Physical Activity and Exercise Index (26), a well-established self-report measure of both lifestyle and structured physical activity in 3 domains, walking, climbing stairs, and exercise/sports. Patients were asked how many blocks they walked and flights of stairs they climbed on average each day during the past week. According to established standards, a block was assigned a value of 8 kilocalories and a flight of stairs was assigned 4 kilocalories (10 stairs was considered 1 flight) (27). Respondents were also asked if they participated in any exercise or sports in the past week, and if so which ones and for how long at each session. Using a standardized compendium of kilocalories for each activity, the composite kilocalories per week from exercise/sports were calculated (28). The kilocalories per week from each domain were calculated separately and an overall total was calculated as the sum of the 3 domains. The sum was then compared with recommended minimum levels, or thresholds, of total weekly energy expenditure, specifically 1,000 and 1,400 kilocalories per week, depending on how activities are performed (18). Energy expenditure from walking was compared with the recommended minimum threshold of 700 kilocalories per week (19).


Multiple covariates have been reported to be associated with greater physical activity. Among the most consistently reported are male sex, more education, more social support, less bodily pain, and fewer depressive symptoms (21, 25). In the current study, social support was assessed with the Duke Social Support Scale (29), which measures the amount of support given by various family and non-family members. An overall score is generated and can range from 0 to 100, with higher scores reflecting more support. Pain was measured by asking participants how much bodily pain they had in the past week with response options ranging from 1 = none to 6 = very severe. Depressive symptoms were measured with the Geriatric Depression Scale (30), a well-established 30-item scale that has been shown to reliably measure psychological symptoms of depression in both older and younger adults. Scores can range from 0 to 30, with higher scores reflecting more depressive symptoms.

Followup at 1 year.

Participants were again interviewed by telephone or in person approximately 1 year later, and asked about current physical activity with the Paffenbarger Index (26). Participants were not given advice on exercise or physical activity during the study.

Statistical analysis.

RA patients and controls were compared according to enrollment characteristics with t-tests and chi-square tests. Paffenbarger Index kilocalories per week were compared between groups in unadjusted, and then adjusted analyses controlling for age, sex, education, social support, bodily pain, and depressive symptoms with multivariate linear regression. Percents of participants in each group meeting recommended energy expenditure thresholds were compared separately with multivariate logistic regression and odds ratios (ORs). For the longitudinal analysis, within-patient changes were calculated as the followup minus the enrollment kilocalories per week, and compared with paired t-tests. Variables associated with more energy expenditure at the followup were identified in multivariate analysis using a stepwise backward elimination process with enrollment characteristics as independent variables and kilocalories per week at the followup as the dependent variable. All analyses were carried out using SAS software (31).

This study was approved by the Institutional Review Board at the Hospital for Special Surgery and the Committee on Human Rights in Research at the Weill Medical College of Cornell University/New York Presbyterian Hospital.


In total, 122 RA patients and 122 controls were enrolled in the larger study comparing workplace events between RA patients and controls (22). Most of these participants (specifically 121 RA patients and 120 controls) also completed all domains of the Paffenbarger Index at enrollment and were included in this analysis. Fifty percent of patients and 58% of controls described the physical demands of their job as light or very light (P = 0.22), and most were in professional occupations (73% and 71%, respectively; P = 0.81). Patients and controls were similar with respect to education, marital status, perceived social support, and depressive symptoms (Table 1). A higher percentage of RA patients were white and had more bodily pain compared with controls. The mean ± SD RA disease duration was 14 ± 10 years, 40% were taking corticosteroids, 49% methotrexate, and 29% etanercept. None of the controls had undergone lower extremity or back orthopedic surgery, as compared with 18% of patients with RA. Foot and/or ankle involvement was reported by 80% of a subgroup of 66 patients with RA.

Table 1. Characteristics of participants
CharacteristicPatients (n = 121)Controls (n = 120)P
  • *

    Based on the Duke Social Support scale (score range 0–100, higher scores = more support).

  • Based on a single question (score range 0–6, higher score = more pain).

  • Based on the Geriatric Depression Scale (score range 0–30, higher score = more depressive symptoms).

Mean age, years (range)49 (19–72)49 (25–70)0.92
Women, %84910.09
Race/ethnicity, %  0.04
 African American1117 
Graduated college, %74700.45
Marital status, %  0.20
 Never married2935 
Social support, mean ± SD*49 ± 1851 ± 160.32
Bodily pain, mean ± SD3.3 ± 1.32.9 ± 1.40.03
Depressive symptoms, mean ± SD6.0 ± 4.95.5 ± 5.30.45

Analysis of enrollment data.

Energy expenditure in kilocalories per week measured by the Paffenbarger Index was compared between groups for each domain and for the overall total (Table 2). Compared with controls, patients with RA had fewer kilocalories per week for the walking domain and for the total category. These differences persisted after adjusting for age, sex, education, pain, social support, and depressive symptoms. In contrast, there were no notable differences between groups for the stair domain and the exercise/sports domain, which had relatively low values for both groups. Thus, the difference between groups for the total category was determined mostly from the walking domain. In subgroup analysis for the patients with RA, there were no differences in energy expenditure according to the use of disease-modifying drugs.

Table 2. Percentage of subjects meeting recommended thresholds of weekly energy expenditure at enrollment according to the Paffenbarger Physical Activity and Exercise Index*
CategoriesPatients (n = 121)Controls (n = 120)Adjusted PAdjusted OR (95% CI)
  • *

    P values adjusted for age, sex, education, pain, social support, depressive symptoms. OR = odds ratio; 95% CI = 95% confidence interval; kcals = kilocalories.

Domains, mean ± SD    
 Walking blocks692 ± 6101,044 ± 1,2600.002
 Stairs184 ± 212185 ± 2620.83
 Exercise/sports599 ± 848729 ± 1,2100.14
 Total1,474 ± 1,1981,958 ± 1,9400.003
Recommended thresholds    
 Total ≥1,000 kcals/week56640.14
 Total ≥1,400 kcals/week41480.17
 Walking ≥700 kcals/week32480.010.48 (0.28–0.84)

Overall, 95% of RA patients and 100% of controls reported walking some blocks, 88% and 86% reported climbing some stairs, and 53% and 49% reported some exercise/sports, respectively. The most common exercise/sports for patients with RA were biking (12%), swimming (10%), aerobics (9%), lifting weights (9%), and running (6%); and the most common for control subjects were lifting weights (13%), running (11%), aerobics (10%), and biking (7%). Compared with those who did not participate in any exercise/sports, those who did participate were more likely in multivariate analysis to be college graduates (P = 0.008; OR 2.2, 95% confidence intervals [95% CI] 1.2–4.0) and to have less pain (P = 0.04; OR 1.2, 95% CI 1.0–1.5).

The percents of patients with RA and controls meeting the thresholds for total weekly energy expenditure were similar (56% and 64% for the lower threshold, and 41% and 48% for the higher threshold, respectively) (Table 2). Thus, although only about half of RA patients met the total weekly recommended total energy expenditure thresholds, their low rates matched the low rates in the healthy controls. However, the percent of RA patients meeting the recommended threshold for walking was notably lower (32%) compared with controls (48%). This difference remained important after adjusting for covariates, with RA patients being markedly less likely to meet the minimum threshold value (Table 2). In subgroup analysis for the RA patients, there were no differences in the percents of patients meeting the recommended thresholds according to the use of disease-modifying drugs. However, RA patients with foot/ankle involvement had less energy expenditure from walking (738 kilocalories per week) compared with RA patients without involvement (870 kilocalories per week). In additional analyses according to demographic characteristics, women had less total energy expenditure than men (1,616 versus 2,385 kilocalories per week; P = 0.01) and among women, more education was associated with more total energy expenditure (P = 0.005 after adjusting for covariates). This was the case for women in both the RA and control groups and was primarily due to more exercise as opposed to more walking or stair climbing. In addition, white patients in the control group had more total weekly energy expenditure compared with other racial/ethnic groups (P = 0.03 after adjusting for covariates) and this difference was primarily due to more walking.

Analysis of followup data.

Of the 121 RA patients, 113 (93%) were contacted for followup at a mean ± SD of 14 ± 1.7 months, and 91 completed the Paffenbarger Index at followup (75% of the initial sample and 81% of those contacted). Of the 120 controls, 109 (91%) were contacted for followup at a mean ± SD of 14 ± 1.7 months, and 87 completed the Paffenbarger Index at followup (73% of the initial sample and 80% of those contacted). Compared with those who were contacted, participants who were not contacted were younger (50 versus 44 years old; P = 0.03) and more likely to be men (8% versus 19%) but did not differ with respect to education, pain, social support, or depressive symptoms. Participants who were contacted for followup but did not complete the Paffenbarger Index were individuals who, for the most part, reported they did not have time available to answer more questions in addition to the employment questions of the larger study. Compared with those who completed the Paffenbarger Index at followup, those who did not were younger (50 versus 45 years old; P = 0.002), but did not differ with respect to sex, education, pain, social support, or depressive symptoms.

To assess the longitudinal pattern of energy expenditure within each group, within-patient differences were calculated as the followup minus the enrollment kilocalories per week for each group separately (Table 3). Among controls, there were slight differences at the 2 time points, however these were not significant for any of the domains or for the total category. Among the patients with RA, there was a decrease in energy expenditure for stairs at followup (P = 0.03), but no differences for the walking and exercise/sports domains or for the total category. We also compared energy expenditure between groups at followup, and found that, similar to the enrollment profile, patients with RA expended fewer kilocalories per week overall (mean ± SD 1,459 ± 1,368 versus 1,928 ± 1,501, P = 0.03) and were less likely to meet the threshold for walking (35% versus 52%; P = 0.03, OR 0.51, 95% CI 0.28–0.92). These results did not differ according to the use of disease-modifying drugs for the RA patients.

Table 3. Energy expenditure of subjects according to the Paffenbarger Physical Activity and Exercise Index at study enrollment and followup
Categories (kcals/week)Patients (n = 91)Controls (n = 87)
EnrollmentFollowupWithin-subject changePEnrollmentFollowupWithin-subject changeP
 Walking blocks691784940.14940978370.64

Finally, all enrollment variables were assessed to determine which variables predicted total energy expenditure at followup. Variables associated with greater total energy expenditure in bivariate analyses were being in the control group (P = 0.03), male sex (P = 0.02), less depressive symptoms (P = 0.03), and greater total energy expenditure at enrollment (P < 0.0001). In multivariate analysis using a backward stepwise elimination process, only greater total energy expenditure at enrollment predicted more energy expenditure at followup (R2 = 0.44, P < 0.0001). The use of disease-modifying drugs at enrollment or at followup did not predict more energy expenditure at followup for RA patients.


This study provides current information on self-reported energy expenditure in patients with RA compared with healthy controls using a well-established scale to measure physical activity. The results showed that RA patients had less weekly total energy expenditure from physical activity than controls, and this was primarily due to less walking and not due to more strenuous activities. Therefore, this study provides evidence that despite the fact that walking is an excellent lifestyle activity, it is markedly underutilized in RA patients. Our study supports the development and implementation of interventions to foster walking as part of a program to address cardiovascular risk in RA patients.

These interventions would differ from traditional exercise programs for RA, which have focused on structured regimens that improve musculoskeletal strength through progressive static and dynamic exercises (32). Although these programs are reported to be safe in terms of clinical and radiologic measures of disease activity (33, 34), rigorous exercise programs or highly structured programs are not widely accepted by patients with RA (11, 35, 36). For example, in a survey to determine patients' attitudes toward exercise, 61% of 606 patients rated lower-intensity exercises just as effective as or better than higher-intensity exercises (11). In another survey of 132 RA patients assessing the role of exercise in managing RA, 30% of patients had unfavorable attitudes toward exercise and 13% stated members of their social network would not approve of them exercising (10, 25). Of the patients who had tried exercises, 73% reported limited to no success in following the regimen due to pain (42%), time constraints (31%), and lack of interest (20%) (10). Other smaller studies showed that motivation and satisfaction with current levels of physical activity vary among RA patients (37) and that women with RA rate themselves low with respect to perceived competence in sports and physical strength (38). In addition, some physicians do not encourage exercise programs for their RA patients. In a survey of 25 rheumatologists, 80% thought certain types of exercise were useful for RA, but only 22% felt confident in instructing patients in exercise and 9% thought their patients would adhere to an exercise regimen (10, 25). Most did not prescribe formal exercises for their patients and few initiated discussions about exercise during routine visits.

Fortunately, recent investigations show that high-intensity activities are not required to improve cardiovascular health, and that moderate-intensity activities also convey notable benefits (13, 39, 40). For the general US population, the Centers for Disease Control and Prevention and the American College of Sports Medicine currently recommend moderate-intensity activities for 30 minutes (bouts of 10 minutes are acceptable) 5 to 7 days per week, which, depending on the way activities are performed, is approximately equivalent to 1,000 to 1,400 kilocalories of energy expenditure per week (16, 17). The American College of Rheumatology recommends physical activity of moderate intensity for 30 to 60 minutes, several times per week for cardiovascular fitness in adults with RA, and calls for the identification of valid measures of physical fitness in RA patients (15).

Fortunately, moderate-intensity activities include many lifestyle activities, such as brisk walking and carrying out daily household chores (20), which are activities that are acceptable to many RA patients. For example, in one study 185 older women with RA completed the Yale Physical Activity Survey and reported that 67% of their weekly activity time was spent doing housework and only 11% was spent in planned exercise (41). Nearly half reported some stair climbing and 67% engaged in weekly leisure walking, which occurred primarily in the context of the daily routine, such as while shopping, walking to public transportation, and running errands.

In the large longitudinal Nurses' Health Study, the 1988 data regarding physical activity showed that women with RA had less weekly energy expenditure from both lifestyle and sports activities compared with women without RA (13.9 versus 15.5 metabolic equivalents per week; P = 0.01) (2). In another study conducted in Sweden, 298 patients with RA were asked to rate the intensity of their daily activities as well as to rate how frequently and for how long they participated in low, moderate, and high- intensity activities during a typical week (42). Overall, 53% of patients met public health recommendations for physical activity, which was the same as or slightly higher than age-matched norms for the general Swedish population.

The results of our study showed that the percent of RA patients who met US threshold values of total energy expenditure was similar to that of controls, with both at low rates of approximately 50%. Our findings parallel those from other studies, which reported that only half of the US population engages in recommended levels of physical activity, and walking is a major method to achieve this goal among patients with all types of arthritis (12, 20). Thus, although the RA patients in our study were comparable with healthy controls, the percents achieving recommended thresholds were disappointingly low, even among RA patients taking disease-modifying drugs. These low rates, however, are potentially of greater clinical importance in RA because the risks of cardiovascular morbidity are higher in RA patients than in the general population (1–3). In addition, it was surprising to find that only 32% of RA patients met the recommended threshold for walking and had markedly less weekly energy expenditure from walking compared with controls. Walking is often considered the ideal lifestyle activity because it is low cost, easily accessible, easily integrated into other activities, and is safe with a relatively low risk of injury (21, 43). Given that our analyses were controlled for pain, it is likely that this low rate was not due to physical limitations, but rather to other factors, such as fear of exacerbating RA, habit, and the mindset of avoiding physical activity (12).

Adopting and maintaining physical activities at levels that promote cardiovascular health are formidable challenges even among healthy individuals. In addition to possible physical limitations, there are multiple demographic and psychosocial barriers to continued physical activity, such as older age, less education, lack of self-efficacy, low social support, lack of time, more depressive symptoms, and not having access to engaging activities (21, 25, 42, 44, 45). Among patients with RA, exercise history has been shown to predict current exercise status (44). For example, one longitudinal study found that RA patients with a history of exercise were 7 times more likely to be exercising 6 months later (25). Our results support these findings with more energy expenditure at baseline predicting greater energy expenditure 1 year later.

There are several limitations to our study. First, all participants were employed individuals in a metropolitan environment, and their physical activity patterns may vary from others in different settings. In addition, all had an established source of medical care where they may have received counseling about exercise. Second, energy expenditure was calculated from self report, thus participants may have over or under estimated their levels of physical activity. Other studies showed that actual energy expenditure is over estimated by 10% and under estimated by 30% when measured with self-report surveys (9). Third, energy expenditure was calculated based on standard estimates of kilocalories required to perform certain activities. It is possible that because of RA some patients may have modified or adapted the way they performed these activities in order to expend less effort, for example by doing activities more slowly. This would have resulted in our over reporting energy expenditure for the RA group. Fourth, comparisons made according to disease-modifying drugs and foot/ankle involvement were based on small subgroups, and therefore are only preliminary results. Further analysis could demonstrate that addressing foot and ankle issues should be an important initial component of a walking program.

Our study is unique because it showed that compared with controls, patients with RA expended less energy in physical activities, primarily due to less walking and not due to more strenuous activities. Walking is a lifestyle activity that is relatively safe in RA and offers patients an excellent opportunity to meet recommended energy expenditure goals. Given the markedly increased risk of cardiovascular disease in patients with RA, fostering prudent physical activity should be a priority in the overall plan to reduce cardiovascular risk. This may be particularly true for patients taking disease-modifying drugs who may continue to be reluctant to exercise because of prior experience with more symptomatic RA. Our results support the development of interventions to foster walking in RA patients with the long-term goal of improving cardiovascular and other health outcomes.


Dr. Mancuso had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Mancuso, Rincon, Paget.

Acquisition of data. Mancuso, Rincon, Sayles.

Analysis and interpretation of data. Mancuso, Sayles, Paget.

Manuscript preparation. Mancuso, Paget.

Statistical analysis. Mancuso.


The authors thank the physicians and patients at the Hospital for Special Surgery and the Cornell Internal Medicine Associates for their participation, and Melanie J. Harrison MD, MS for her assistance with the Hospital for Special Surgery Rheumatoid Arthritis Clinical Outcomes Registry.